The other day I found myself lost in the thought about how QA-s in the gaming industry perform the testing right before an important launch or an update.
Mind you, I’m talking about PC/Console games at this point, lets leave mobile gaming out of the question for now since the complexity is probably a lot lower compared to PC and consoles.
So I tried to apply my own experience on automating and testing from frontend and backend API testing but I couldn’t wrap my head around how QA-s in gaming plan and execute tests.
To be more specific, lets think about a new patch for, say, Diablo 4 - Imagine the new patch brings a lot of new spells and animations to certain playable characters.
I can’t fathom how do they test all the probable test scenarios related to the addition of even 1 spell in the game - I would guess that QA’s have to check all the probable scenarios where that spell can be used (for example: how the spell combines with tons of other already-existing spells in the game, how the spell works in different planes and environments of the game, how the spell damage/healing or other specifics are calculated against enemy defense rating/resistances, etc, etc. The list goes on!)
If there’s anyone who had touchpoints with QA in gaming, I would love to hear a bit about how the testing is done in this industry and what tools are commonly used in the process.
Not many games testers here, it’s often not paying at the pro level. I once went through a Sony interview, they have all kinds of tooling for testing, it’s very much engine specific stuff that gets tested and how it gets tested is not always in ways players expect. The tooling is where the buck are as far I could see.
See Game Testing 2023 . It does not really work like that IRL. Releases are staggered and publishers rely on free testing they get from beta testers who sign an NDA in return for early drops. And mechanically, in house testers use all kinds of tooling to help them set up the scenarios and actually don’t end up doing such crazy analysis at the user exposed surfaces, the engine itself is where the logic lies, it’s easier to test there, not by actually playing.
Been a bit since I worked in the sector and it will vary massively by game & studio but:
There is usually some E2E / UI level automation. This can vary a lot by the type of game, platform etc.
There’s usually tooling to help with load testing etc, similar to with web applications etc.
Manual testing is massively important. It isn’t quite end-user level. For example you might be in a test room with debug commands to spawn objects on demand then interact with them as a normal user.
Larger games will have fairly sizeable QA teams who are testing it. There may also be embedded testers.
Larger publishers will have additional QA teams, especially when you’ve got compatibility, languages etc.
From my experience there wasn’t a huge amount of planning involved in our testing. It was often adhoc, responding to changes going into builds… although our engineering practices were pretty rubbish.
I have been 11 years in gamedev.
Tests depend on the type of game. There are types of games, which don’t need testing.
About the patch of Diablo 4, In this company, there are problems with the company.
Verifications defence on priority, time and opportunity.
Without context, It is difficult to explain.
(It was ~2006 I visited Sony in London. It was bizarre, pizza boxes, soda machines, a floor full of cubicles and everyone playing hard. Horrible office location, but lots of manual testing in-house does happen.) Glad I never got sucked in, but also sad.
Welcome to the community @scittels , we would love to hear more, not the politics, but about the stack and methodology.
